To the Editor—Central venous catheter (CVC)–related bloodstream infections (CRBSI) are associated with high morbidity, especially in neutropenic patients.Reference Hentrich, Schalk and Schmidt-Hieber ¹ In general, obesity is one of the risk factors associated with CRBSI. However, this finding is inconsistent in the literature and no data are available on patients with hematologic malignancies so far. CVC insertion is often technically challenging in obese patients owing to obscured landmarks of the neck.Reference Graham, Ozment, Tegtmeyer, Lai and Braner ^{2 ,} Reference McGee and Gould ³ Furthermore, with increased sweating,Reference Lawrence and Kopelman ⁴ bandages at the CVC insertion site are frequently problematic and lead to higher risk of CRBSI.Reference Trick, Miranda, Evans, Charles-Damte, Reilly and Clarke ⁵ Therefore, we hypothesized that obese hematologic patients treated with chemotherapy are at higher risk of having CRBSI than non-obese hematologic patients.

In the monocentric, prospective SECRECY registry (Study to Evaluate Central Venous Catheter-Related Infections in Hematology and Oncology; German Clinical Trial Register number, DRKS00006551) we evaluated CRBSI in all consecutive patients in our department from March 20, 2013, through March 13, 2015. The registry was approved by the local ethics committee (approval no. 84/14). Owing to anonymization of patients’ data, written informed consent was not required within the registry. However, written informed consent is available for the procedure of CVC insertion during routine clinical use. All patients were at least 18 years of age. CVCs were mainly used for chemotherapy or best supportive care. All CVCs were inserted under sterile conditions and ultrasound guidance by experienced physicians. To define CRBSI we used the 2012 criteria of the Infectious Diseases Working Party of the German Society of Hematology and Medical Oncology.Reference Hentrich, Schalk and Schmidt-Hieber ¹ Obesity was defined as body mass index (BMI; calculated as weight in kilograms divided by height in meters squared) of 30 or greater according to the commonly known World Health Organization criteria.

Altogether we analyzed 335 triple-lumen, short-time CVCs (including 11 coated CVCs) that were in use at least 1 day. The CVCs accounted for a total of 5,094 CVC days (mean [range], 15.2 [1–60] days). The mean (range) age of the 176 patients was 58.1 (25–81) years; 106 (60.2%) were men. Mostly, patients had acute leukemia (178 [53.1%]), multiple myeloma (83 [24.8%]), or malignant lymphoma (41 [12.2%]). Patients received a mean (range) of 1.9 (1–6) CVC insertions within the observation period. Most CVCs (314 [93.7%]) were inserted into the internal jugular vein. CRBSI were detected in 77 cases (12 [15.6%] definite, 21 [27.3%] probable, 44 [57.1%] possible), with an incidence of 15.1/1,000 CVC-days and an incidence rate of 23.0%. The most prevalent pathogen was Staphylococcus epidermidis (43 [71.7%]). At baseline—that is, at time of the first CVC insertion of the patients, the mean (range) BMI was 28.0 (15–46). Taking all 335 CVCs together, the BMI of the patients was a mean of 27.3, whereas in 95 CVCs (28.4%), the BMI of the patients was at least 30 (for the subgroup of the obese patients the mean BMI was 33.8).

Complications of CVC insertion (bleeding, hematoma, >2 punctures, or malpositioning of the guidewire) were reported in 18.3% of obese patients and in 18.4% of non-obese patients. This indicates no increased risk for complications during CVC insertion among obese patients (odds ratio [OR], 0.99].

Comparing the CRBSI rate in obese and in non-obese patients we found no differences in CRBSI frequency (22.1% vs 23.3%; OR, 0.93).

Duration of CVC use appeared to be significantly shorter in obese compared with non-obese patients (13.5 vs 15.9 days). However, using the modified Infection Probability Score,Reference Schalk, Hanus, Färber, Fischer and Heidel ⁶ which is more stringent in defining neutropenia than the original IPS,Reference Peres Bota, Mélot, Lopes Ferreira and Vincent ⁷ we found a higher modified Infection Probability Score at the time of CVC insertion in obese than in non-obese patients (7.6 vs 5.8). Interestingly, sex is not a risk factor for CRBSI in obese patients (men vs women, OR, 0.86 [95% CI, 0.32–2.35]; P=.97). CRBSI risk was increased neither for obese men (OR, 0.68) nor for obese women (OR, 1.58) (data are summarized in Table 1).

TABLE 1 Data on 176 Patients, Use of 335 Central Venous Catheters (CVCs), and Occurrence of Catheter-Related Bloodstream Infection (CRBSI)

Note. Body mass index (BMI) is calculated as weight in kilograms divided by height in meters squared. Obesity is defined as BMI≥30. mIPS, modified Infection Probability Score; OR, odds ratio.

^a Fisher exact test.

^b Student t test.

In our experience, CVC insertion with support of ultrasonography is a safe procedure in obese hematologic patients. Surprisingly, obesity could not be defined as a risk factor for CRBSI in our dataset. Using the modified Infection Probability ScoreReference Schalk, Hanus, Färber, Fischer and Heidel ⁶ as a tool to describe the grade of illness of patients (with all critical parameters such as body temperature, heart rate, respiratory rate, absolute neutrophil count, and C-reactive protein as well as the Sequential Organ Failure Assessment scoreReference Vincent, Moreno and Takala ⁸ being included), we found that obese patients had increased values and were therefore more challenged by the disease at the time of CVC insertion; however, the risk for CRBSI per se was not increased. One potential pitfall that could mask the CRBSI risk in obese patients is the duration of CVC use, which was significantly shorter (2–3 days shorter) in obese vs non-obese patients in the cohort investigated. CRBSI are known to be associated with the duration of CVC use.Reference Schalk, Hanus, Färber, Fischer and Heidel ^{6 ,} Reference Pepin, Thom and Sorkin ⁹

Another piece of data supporting our findings is the inverse association between obesity and risk of febrile neutropenia that has been reported recently.Reference Chao, Page, Yang, Rodriguez, Huynh and Chia ¹⁰ Potential mechanisms include altered pharmacokinetics and/or reduced relative efficacy of chemotherapy due to obesity.Reference Chao, Page, Yang, Rodriguez, Huynh and Chia ¹⁰